INDEX

Explanations

phrases emphasizing comparisons and equivalence

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 14-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_14/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.14.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 âgées

-0.57

Abp

-0.41

 médié

-0.40

 ktorá

-0.40

 âgé

-0.40

zelfde

-0.39

 ilustracja

-0.37

 Polri

-0.37

 étrangère

-0.37

 colgante

-0.36

POSITIVE LOGITS

 ویکی‌پدی

0.65

kool

0.56

Hal

0.56

!):

0.54

squ

0.53

Hal

0.53

suicide

0.53

 wiſſen

0.53

Cic

0.52

••••

0.52

Activations Density 0.118%

phrases emphasizing comparisons and equivalence

No Comments

No Known Activations